Research Article
Advances in Mechanical Engineering
2017, Vol. 9(7) 1–12
Three-dimensional air distribution  The Author(s) 2017DOI: 10.1177/1687814017707414
journals.sagepub.com/home/ade
analysis of different outflow typed
operating rooms at different inlet
velocities and room temperatures
Hande Ufat, Omer Kaynakli, Nurettin Yamankaradeniz and
Recep Yamankaradeniz
Abstract
It is important to provide a regular unidirectional air distribution in an operating room to reduce the number of parti-
cles. Measurements were taken in the one of the operating rooms at Uludag University Medical School, with laminar air
flow unit and two-cornered outlet which was thought to have some airflow problems. Moreover, a three-dimensional
computational fluid dynamics model has been developed where the measurements were taken in. The distributions of
air velocity, temperature, and relative humidity have been examined and compared with the measurements to validate
the computational fluid dynamics analyses. In addition to present model, four-cornered outlet operating room has been
analyzed and compared with the results of two-cornered one. It is concluded that the case of four-cornered outlet pro-
vides more suitable thermal distribution, which results in a reduction of the particle numbers in the interior. Although
there is no significant change in temperature and relative humidity in the operating room, air distribution changes
dramatically.
Keywords
Operating room air conditioning, computational fluid dynamics, air distribution, laminar air flow unit, validation, numeri-
cal simulation
Date received: 8 December 2016; accepted: 6 April 2017
Academic Editor: Oronzio Manca
5
Introduction ASHRAE, DIN 1946/4, and VDI 2167; Table 1
defines the environmental parameters of these
An operating room (OR) air conditioning system is standards.
designed to reduce the risk of infection to the patients Turbulent air flow was allowed in previous stan-
and OR staff and also provide the best possible thermal dards and guidelines but laminar air flow (LAF)
comfort.1,2 Poorly ventilated ORs not only make the
surgical staff and the patient uncomfortable but also
can spread infection.3 If pathogenic particles enter a
Department of Mechanical Engineering, Faculty of Engineering, Uludag
surgical wound, the risk of infection at the surgical site University, Bursa, Turkey
increases.4 Air distribution plays very important role to
success this condition. Corresponding author:
Although there are standards and guidelines for Hande Ufat, Department of Mechanical Engineering, Faculty of
Engineering, Uludag University, Gorukle Campus, TR-16059 Bursa,
almost every country, some of them are preferred. Turkey.
These preferred standards and guidelines are Email: handet@uludag.edu.tr
Creative Commons CC-BY: This article is distributed under the terms of the Creative Commons Attribution 4.0 License
(http://www.creativecommons.org/licenses/by/4.0/) which permits any use, reproduction and distribution of the work without
further permission provided the original work is attributed as specified on the SAGE and Open Access pages (https://us.sagepub.com/en-us/nam/
open-access-at-sage).
2 Advances in Mechanical Engineering
Table 1. International technical HVAC guidelines that define the environmental parameters in an OR.4
Room air temperature Tr (C) Supply air Relative Supply air velocity (m/s) Standards and guidelines
temperature Ti (C) humidity (%)
19–26 Ti\Tr a v 0.23 DIN 1946
22 Ti\Tr 30–50 v 0.23 VDI 2167
20–24 a 30–60 0.13–0.18 ASHRAE 170
17–27 a 45–55 1.3–1.8 ASHRAE application handbook
aValue not specified.
systems are required in the recently published stan- effect of decreasing the supply air velocity and different
dards. Even though different supply air velocities are lamp positions. They found that different supply air
recommended according to Table 1 and other stan- velocities and lamp positions have slight effects on the
dards, a velocity of more than 0.45m/s will cause draft thermal comfort but serious effects on the movement of
discomforts for the surgical staff.4 infectious particles. They also performed numerical cal-
Many experimental studies and computational fluid culations. An increasing trend in the bacteria carrying
dynamics (CFD) analyses have been presented about particle (BCP) concentration was observed as the num-
thermal comfort and infection factors in ORs. Chow ber of staff in the OR increased. Sadrizadeh and
and Yang6–8 explained the basic principles of an OR Holmberg16 investigated the effects of a mobile LAF
and applied the CFD method in their investigations. unit on the concentration, deposition, and distribution
They studied seven different cases with different medi- of BCPs in an OR and found that the airborne bacteria
cal lamp positions and different supply air velocities, were decreased by adding a mobile LAF unit. El
with or without partial walls. They stressed the health Gharbi et al.17 used experimental measurements and
risk of the airborne bacteria released from both the sur- mathematical modeling techniques to determine the
gical staff and the patient. Ho et al.3 presented a three- thermal and relative humidity characteristics in an OR.
dimensional CFD analysis for thermal conditions and They examined different variations of supply and
contaminant removal in an OR. They investigated the exhaust air locations and found that the optimum way
effects of horizontal locations of supply and exhaust to sterilize the operating area was using unidirectional
grilles on thermal comfort and contaminant removal. air supply. Woloszyn et al.18 studied numerical simula-
Rui et al.9 studied the airborne transmission of bacteria tion and experimental measurements of the airflow pat-
in two ORs during surgery. They also conducted a terns and the contaminant distribution in an OR with a
numerical simulation to calculate the particle trajec- diagonal air distribution system.
tories. Their results showed that improving of the air Sadrizadeh et al.19 investigated the effectiveness of
flow pattern can reduce the particle deposition on criti- vertical and horizontal ventilation systems in terms of
cal surfaces; however, the effect is less clear when reducing sedimentation and distribution of bacteria
increasing the air change rate by a certain amount. carrying particles in an OR. Their results show that a
They also found that bacteria colony deposition laminar and well-organized flow pattern is retired for a
increased if the air velocity increased above a particular good result.19 Sadrizadeh and Holmberg20 investigated
velocity. Balaras et al.10 presented an overview of the the effects of a mobile LAF unit on the concentration,
general design of indoor air conditions in an OR. They deposition, and distribution of BCPs in an OR and
investigated 20 ORs at 10 different hospitals including found that the airborne bacteria were decreased by
an audit that recorded information on construction, adding a mobile LAF unit. Liu et al.21 investigated an
indoor air conditions, and type of heating, ventilating, alternative horizontal airflow and also evaluated the
and air conditioning (HVAC) system. Van Gaever effectiveness of the horizontal unidirectional airflow to
et al.4 summarized the technical standards and com- control infectious airborne particles through on-site test
pared them to the thermal comfort standard. Dharan and CFD simulation design. Nastase et al.22 reviewed
and Pittet11 stated that a decrease of particle numbers the main European standards on design indoor thermal
in an OR reduces the surgical site infection for orthope- conditions and operation of HVAC systems for ORs.
dic surgery. Memarzadeh andManning12 studied differ- Kobayashi et al.23 developed a calculation model to
ent ORs with different ventilation systems. Their results predict vertical temperature profile in an impinging jet
show that laminar flow systems are the best choice for ventilation (IJV) system. They presented a parametric
ORs. Pasquarella et al.13 and Smith et al.14 studied effi- study on the supply air velocity is performed based on
ciency of LAF units in decreasing bacterial contamina- setting and its effect on the thermal stratification.
tion and found a positive correlation between two In this study, two ORs, one containing four (two
variables. Chow et al.15 investigated the integrated near the flour and two near the ceiling) outlet grilles in
Ufat et al. 3
two corners and one containing eight (four near the The generation of turbulence due to buoyancy is
ceiling and four near the floor) outlet grilles in four cor- given by Gb ¼ bgiðmt=PrtÞð∂T=∂xiÞ, where Prt is the tur-
ners, were modeled and analyzed in CFD program, bulent Prandtl number for energy and gi is the compo-
upon which air flow, temperature, and relative humid- nent of the gravitational vector in the ith direction. For
ity distribution have been examined. the standard k–e model, the default value of Prt is 0.85.
The coefficient of thermal expansion, b, is defined as
b ¼ ð1=rÞð∂r=∂TÞp. The dilatation dissipation termMaterial and method YM is modeled as YM ¼ 2reM2t , where Mt is the turbu-
Theoretical formulation lent Mach numpbeffirffiffi;ffiffiffiffitffiffihe turbulent Mach number is
defined as Mt ¼ k=a2 where a ([gRT) is the velocity
The software ANSYS 14.5 Fluent was used for the of sound.
CFD analysis. The standard k–e model and the stan- The discretized method selected for the convection
dard wall function were used as the computational terms was the second-order upwind scheme. The
tools. The simulation was conducted in steady-state SIMPLE algorithm was chosen to couple pressure and
condition. The transport equations of the standard k–e velocity. The Boussinesq model was used to consider
turbulence model can be written as follows:24 the buoyancy effect. The convergence criteria were set
to 1 3 1025 for all equations except energy, for which it
Conservation of mass equation was 1 3 1026. The turbulent intensity and hydraulic dia-
meter were 5% and 3m, respectively. The distribution of
rðr~vÞ ¼ Sm ð1Þ the air exhaust was approximately 25% from the outlet
diffusers near the ceiling and 75% from the outlet diffusers
Momentum
  near the floor. Outlet type was chosen as ‘‘Outflow,’’ andflow rate weighting was set according to this information.
r  ðr~v~vÞ ¼ rp+r  t + r~g ð2Þ
where t is viscous stress and can be written as CFD model and meshing
t ¼ m½ðr~v+r~vT Þ  ð2=3Þr ~vI . The modeling of the two above-mentioned ORs was
conducted using ANSYS 14.5 Workbench program,
Energy conservation while numerical analysis was conducted using Fluent
 X ! program. In order to confirm the accuracy of the analy-sis, the measurements were taken from the OR with
r  ð~vðr~v+ pÞÞ ¼ r  hjJj + Sh ð3Þ outlet air provided from two corners and then com-
j
pared to the results from the Fluent program. This OR,
Turbulent kinetic energy one of the ORs in Uludağ University Medical Faculty
   Hospital, is shown in Figure 1.
∂ ð Þ ¼ ∂ mt ∂krkui m+
∂xi ∂xj sk ∂xj
+Gk +Gb  re  YM + Sk
Dissipation rate of turbulent kinetic energy
  
∂ ð Þ ¼ ∂ m ∂ereui m+ t
∂xi ∂xj se ∂xj ð4Þ
e e2
+C1e ðGk +G3eGbÞ  C2er + Se
k k
The turbulent viscosity was computed using
mt ¼ rC ðk2m =eÞ. Cm, C1e, C2e, sk, and se are constants
and their values are Cm=0.09, C1e=1.44, C2e=1.92,
sk=1.0, and se=1.3. The term Gk, representing the
production of kinetic energy, can be defined as
Gk ¼ ruiujð∂uj=∂xiÞ. To evaluate this parameter in a
manner consistent with the Boussinesq hypothesis, it is
used Gk ¼ m 2tS , where the modpulffiuffiffisffiffiffiffioffiffiffifffiffi the mean rate- Figure 1. The operating room where the OR measurements
of-strain tensor is defined as S[ 2SijSij. were taken in.
4 Advances in Mechanical Engineering
Figure 4. Mesh independency.
Figure 2. CFD model of the operating room.
Validation
The air velocity, temperature, and relative humidity in
the room were measured in six points of surface of the
LAF unit, the average value was taken, and the average
air inlet conditions were determined. Also, the air velo-
city was measured in spots located on a specific range
from each other from the ceiling to the floor. Besides,
the temperature and relative humidity in the area below
LAF unit were measured, the average value was taken,
and the room temperature and relative humidity were
determined. Room temperature (To) was calibrated
with reference to the standardized room temperature
between 19C and 22C with 1 increase and inlet air
velocity average was measured as 0.2m/s. Velocity–
height diagrams drawn using the conducted velocity
Figure 3. Mesh model of the operating room.
measurements and the values obtained from the analy-
sis are shown in Figures 5–8. As it can be observed from
the charts, the room measurements results and CFD
OR inlet air is provided through 3.2m 3 3.2m analysis results are rather close; thus, the analysis is
sized LAF unit. Outlet air is provided through two verified.
0.525m 3 0.525m sized outlet ports near the floor Recommendation for obtaining the optimal air dis-
and two 0.425m 3 0.325m sized outlet ports near tribution in ORs is to place eight outlet grilles in four
the ceiling. The locations of LAF units set above two corners; 2/3 of the air should be absorbed from the
OR lamps are shown in Figure 1. The lamps were low-level grilles, while 1/3 of the air should be absorbed
modeled in this way because the OR personnel pre- from the high-level grilles.10 In order to observe how
fers them in this location. air velocity, temperature, and relative humidity will
Modeled ORs are shown in Figure 2, while the behave in the room measured and designed in above-
mesh model is shown in Figure 3. The analysis was mentioned way, the same room was modeled with out-
conducted with different mesh numbers for room let air installed in four corners and the CFD analysis of
with outlet air provided from two corners in order to each room was conducted on To=18C–24C with 1
test mesh independence and the obtained air increase. The analysis was conducted by keeping the
velocity–height chart is shown in Figure 4. The anal- temperature of the inlet air at 3C lower than the room
yses were conducted with 176,069; 248,562; 287,310; temperature. The temperature needs to be 2C–3C
316,591; 416,471 and 532,420 mesh numbers and it lower than the room temperature in order for air to
was observed that the air velocity related to the last descend curtain-like. In these conditions, the air density
three mesh numbers was almost identical. The analy- will be higher than the room air density; thus, with the
sis was conducted with 316 591 mesh number in effect of gravity, the air will be pulled down by its own
order to shorten the CFD analysis conducting weight. Also, CFD analysis was conducted in the same
period. temperature conditions accompanied by 0.25 and
Ufat et al. 5
Figure 5. Velocity–height graph for Tr = 19C and Vi = 0.2 m/s. Figure 7. Velocity–height graph for Tr = 21C and Vi = 0.2 m/s.
Figure 6. Velocity–height graph for Tr = 20C and Vi = 0.2 m/s. Figure 8. Velocity–height graph for Tr = 22C and Vi = 0.2 m/s.
0.35m/s inlet air velocities and the results were better on low temperatures when the inlet air velocity is
crosschecked. 0.2m/s.
The air velocity distribution if Vg=0.25m/s with
section x=2.8m is shown in Figure 10. In this velocity,
Results and discussion the velocity distribution resembles the above mentioned
The numerical analysis of an OR with outlet grilles Vg=0.2m/s. With the increase in inlet air velocity, the
outlet air velocity increases as well and the air is directed
collecting air from two corners toward the outlet grilles without descending sufficiently
The air velocity distribution if Vg=0.2m/s with section toward the floor. Particularly, with the increasing tem-
x=2.8m in an OR with air outlet provided from two perature, it is observed that there is an air movement
corners is shown in Figure 9. At first glance, it may around the OR table and toward outlet. These air move-
seem that distributions resemble each other. However, ments can cause an increase in the number of particles.
even though inlet air velocity is the same for each anal- The air velocity distribution if Vg=0.35m/s with sec-
ysis, as the room temperature increases and the air den- tion x=2.8m is shown in Figure 11. When this inlet velo-
sity decreases, the air descending velocity decreases and city is taken into consideration, it can be observed that the
the air does not fulfill its curtaining purpose. Buoyancy room temperature does not significantly effect the air dis-
effect is efficient so air cannot overcome the thermal tribution. The velocity distribution is the same on every
plumes and unidirectional flow was disrupted. High air room temperature. Also, since the outlet velocity increased
velocity causes turbulence and disturbs the particles excessively, the air is quickly directed toward the outlet
in the operating area, thus increasing the risk of infec- grilles. Air velocity around the table is high and it can
tions. It can be said that the air curtain effect is formed cause the increase in the particles number.
6 Advances in Mechanical Engineering
Figure 9. Velocity distribution for at x = 2.8 m for Vinlet = 0.2 m/s.
The numerical analysis of an OR with outlet grilles
collecting air from four corners the air from being directed to the outlet grilles to fast.
Thus, the air descends toward the floor in a more regu-
Velocity distributions for Vg=0.2m/s on different lar fashion and forms a curtain. Since the air velocity is
room temperatures are shown in Figure 12. If this air lower around the operating table, there is no turbulence
distribution in this system has been examined, it is and the particles are less dispersed comparing to the
observed that room temperature does not effect des- other system. Thermal plumes cannot disrupt the uni-
cending velocity of the air as much as in the other sys- directional flow pattern, so the analyses are nearly simi-
tem. Since the inlet air is divided on eight grilles, the lar between 18C and 24C. When the area outside the
outlet air velocity is significantly lower. This obstructs operating area is examined, it can be observed that the
Ufat et al. 7
Figure 10. Velocity distribution for at x = 2.8 m for Vinlet = 0.25 m/s.
air in this part to is more stagnant comparing to the air velocity outside the operating area increases in
the other system. Thus, the number of particles dis- relation to the outlet velocity. This velocity is signifi-
persed from this area to the operating area will be sig- cantly lower comparing to the system where air outlet
nificantly decreased. is conducted from two corners. It is easily observable
The analyses conducted for the same system for from the analyses conducted on both systems that the
Vg=0.25m/s and Vg=0.35m/s are shown in lamps are breaking the air curtain. Since the air below
Figures 13 and 14, respectively. The velocity profile the lamps is stagnant and the sweeping of particles in
below the LAF unit resembles Vg=0.2m/s velocity, this area is obstructed, the number of particles in this
but it was observed that the velocity increased in rela- area may increase. The air curtain can be properly
tion to the inlet velocity. When the velocity increases, formed in the system where air outlet is conducted
8 Advances in Mechanical Engineering
Figure 11. Velocity distribution for at x = 2.8 m for Vinlet = 0.35 m/s.
from four corners with Vg=0.35m/s and the air stag- with section x=2.8m for To=20C and Vg=0.35m/
nancy below the lamps can be prevented. s is shown in Figure 15, while the relative humidity dis-
The air temperature and relative humidity in the tribution for To=20C and Vg=0.35m/s is shown in
room do not differ much depending on the system. The Figure 16.
temperature and relative humidity distribution are
nearly same at the measured temperature and inlet velo-
city in this study. As there is not much heat loss in the Conclusion
OR, the temperature and relative humidity value in the According to the obtained findings, the results were sig-
OR approaching the inlet conditions in a short time. nificantly better when eight outlet grilles, set in such
For this reason, complete analysis results were not pre- way that 2/3 of the air is absorbed from the low-level
sented. The temperature distribution for both systems grilles, while 1/3 is absorbed from the high-level grilles,
Ufat et al. 9
Figure 12. Velocity distribution for at x = 2.8 m for Vinlet = 0.2 m/s.
were providing air flow from four corners comparing has been determined that velocity above 0.45m/s is
to the same operation conducted by four outlet grilles uncomfortable (Gaever et al.). The inlet air velocity
in two corners. Due to the large amounts of inlet air increases in the systems with outlet grilles installed in
related to the LAF unit size, outlet air velocity is high two corners, thus preventing proper air distribution
when four grilles are installed and inlet air is directed around operating table. With this system, the inlet air
to the exit without forming the air curtain. For the velocity should not be too high in order to prevent tur-
same reason, the velocities outside the OR area are bulence inside the room.
quite high. The recommended inlet air velocity in the As mentioned above, the obtained air distribution is
ORs with LAF unit system is Vg 0.23m/s; however, it more regular if the system has outlet provided from
10 Advances in Mechanical Engineering
Figure 13. Velocity distribution for at x = 2.8 m for Vinlet = 0.25 m/s.
four corners. Additionally, room temperature does not The effect of the two examined systems on the tem-
significantly effect the distribution when inlet air velocity perature and relative humidity was not significant.
is low. Besides significantly better air distribution provided However, installing outlet grilles in four corners to pro-
within the OR area, the air velocity outside the operating vide proper air distribution inside the room is more
area is lower comparing to the other system, thus decreas- appropriate since the main goal is to decrease the risk
ing the particle dispersion inside the room. If the inlet air of an infection inside the room. Also, it has been
velocity is kept around Vg=0.25m/s for comfort reasons, demonstrated that increasing the inlet air velocity in
the forming of dead zone above operating table could be the system with outlet grilles in two corners is not an
prevented by changing the lamp positions. appropriate solution.
Ufat et al. 11
Figure 14. Velocity distribution for at x = 2.8 m for Vinlet = 0.35 m/s.
Figure 15. Temperature distribution at x = 2.8 m for To = 20C and Vg = 0.35 m/s.
12 Advances in Mechanical Engineering
Figure 16. Relative humidity distribution at x = 2.8 m for To = 20C and Vg = 0.35 m/s.
Declaration of conflicting interests 10. Balaras CA, Dascalaki E and Gaglia A. HVAC and
The author(s) declared no potential conflicts of interest with indoor thermal conditions in hospital operating rooms.
respect to the research, authorship, and/or publication of this Energ Buildings 2007; 39: 454–470.
article. 11. Dharan S and Pittet D. Environmental controls in operat-
ing theatres. J Hosp Infect 2002; 51: 79–84.
Funding 12. Memarzadeh F and Manning AP. Comparison of operat-
ing room ventilation systems in the protection of the sur-
The author(s) disclosed receipt of the following financial sup- gical site/discussion. ASHRAE Tran 2002; 108: 3.
port for the research, authorship, and/or publication of this 13. Pasquarella C, Sansebastiano GE, Ferretti S, et al. A
article: This work was supported by The Scientific Research mobile laminar airflow unit to reduce air bacterial con-
Project Department of Uludag University under the Project tamination at surgical area in a conventionally ventilated
No. OUAP (M) 2013/2. operating theatre. J Hosp Infect 2007; 66: 313–319.
14. Smith EB, Raphael IJ, Maltenfort MG, et al. The effect
of laminar air flow and door openings on operating room
References contamination. J Arthroplasty 2013; 28: 1482–1485.
1. Vilain R, Pereira ML, Felix V, et al. Thermal comfort and 15. Chow TT, Lin Z and Bai W. The integrated effect of
local discomfort in an operating room ventilated with medical lamp position and diffuser discharge velocity on
spiral diffuser jet. HVAC&R Res 2013; 19: 1016–1022. ultra-clean ventilation performance in an operating the-
2. Wang FJ, Lai CM, Cheng TJ, et al. Performance evalua- atre. Indoor Built Environ 2006; 15: 315–331.
tion of indoor environment parameters for an unoccu- 16. Sadrizadeh S and Holmberg S. Effect of a portable ultra-
pied operating room. ASHRAE Tran 2011; 117: 557. clean exponential airflow unit on the particle distribution
3. Ho SH, Rosario L and Rahman MM. Three-dimen- in an operating room. Particuology 2015; 18: 170–178.
sional analysis for hospital operating room thermal 17. El Gharbi N, Benzaoui A, Khalil EE, et al. Analysis of indoor
comfort and contaminant removal. Appl Therm Eng air quality in surgical operating rooms using experimental and
2009; 29: 2080–2092. numerical investigations.Mech Ind 2012; 13: 123–126.
4. Van Gaever R, Jacobs VA, Diltoer M, et al. Thermal 18. Woloszyn M, Virgone J and Mélen S. Diagonal air-
comfort of the surgical staff in the operating room. Build distribution system for operating rooms: experiment and
Environ 2014; 81: 37–41. modeling. Build Environ 2004; 39: 1171–1178.
5. Anıl OB, Mobedi M and Ozerdem MB. Hastane hijyenik 19. Sadrizadeh S, Holmberg S and Tammelin A. A numerical
ortamları ixcin klima ve havalandırma sistemleri tasarım investigation of vertical and horizontal laminar airflow venti-
parametreleri. In: Proceedings of the 8th national sani- lation in an operating room. Build Environ 2014; 82: 517–525.
tary engineering congress, I_zmir, 25–28 October 2007, 20. Sadrizadeh S and Holmberg S. Effect of a portable ultra-
pp.497–509. clean exponential airflow unit on the particle distribution
6. Chow TT and Yang XY. Performance of ventilation sys- in an operating room. Particuology 2015; 18: 170–178.
tem in a non-standard operating room. Build Environ 21. Liu J, Wang H and Wen W. Numerical simulation on a
2003; 38: 1401–1411. horizontal airflow for airborne particles control in hospi-
7. Chow TT and Yang XY. Ventilation performance in tal operating room. Build Environ 2009; 44: 2284–2289.
operating theatres against airborne infection: review of 22. Nastase I, Croitoru C, Vartires A, et al. Indoor environ-
research activities and practical guidance. J Hosp Infect mental quality in operating rooms: an European stan-
2004; 56: 85–92. dards review with regard to Romanian guidelines. Energy
8. Chow TT and Yang XY. Ventilation performance in the Proced 2016; 85: 375–382.
operating theatre against airborne infection: numerical study 23. Kobayashi T, Sugita K, Umemiya N, et al. Numerical
on an ultra-clean system. J Hosp Infect 2005; 59: 138–147. investigation and accuracy verification of indoor environ-
9. Rui Z, Guangbei T and Jihong L. Study on biological ment for an impinging jet ventilated room using computa-
contaminant control strategies under different ventilation tional fluid dynamics. Build Environ 2017; 115: 251–268.
models in hospital operating room. Build Environ 2008; 24. ANSYS. ANSYS fluent theory guide. Canonsburg, PA:
43: 793–803. ANSYS, 2015.